Вы находитесь на странице: 1из 51

The Definitive Guide to Position Sizing Strategies

The following includes two excerpts from the 1st Edition of the Definitive Guide to Position Sizing
Strategies. Although these sections are not included in the new edition of the book, we still believe the
information to be very useful and so we have decided to include it as a bonus with the 2nd Edition. Some
of the more significant information contained here can still be found in other parts of the 2nd Edition of
the Definitive Guide to Position Sizing Strategies, but these particular sections were removed in full.

Part I: Are You Doomed to Failure? 1

Judgmental Shortcuts 1

Bias 1: Locus of Control—The Lotto Bias 2

Bias 2: The Need to Be Right 4

Bias 3: Percent Gain 8

Bias 4: Lots of Input Says the Same Thing 10

Bias 5: Authority 11

Bias 6: Prediction and Understanding 12

Bias 7: Wanting Lots of Facts 15

Other Biases That Influence Being Right 16

Bias 8: The Law of Small Numbers 16

Bias 9: Once We Think We’ve Got It, It’s Hard to Get Rid of It 17

Bias 10: Representation 18

Conclusion 20

Part II: Position SizingSM Software Examined 21

My Experiences with Position Sizing Software 21

Software to Keep Track of Your Trades 23

Simulation Software 30

Position Sizing Software 33

System Specific Software with Position Sizing Capabilities 34

Multi-Purpose Software that Includes Position Sizing 36

High-End Software 46

Conclusion 49
The Definitive Guide to Position Sizing Strategies

Part I:

Are You Doomed to Failure?

Despite the importance of the material presented so far in Part I, most people have psychological
biases that will cause them to 1) ignore the material totally or 2) do exactly the opposite of what is
recommended. As a result, in this chapter, I want to show you some of those biases and what you
can do to overcome them.

Judgmental Shortcuts
Why Judgmental Shortcuts Are Important: French Economist George Anderla found that the
rate of information flow with which we human beings must cope doubled in the 1,500 years
between the time of Jesus and Leonardo DaVinci. By the year 1750 (i.e., in about 250 years), it
doubled again. The next doubling only took about 150 years to about 1900. The onset of the
computer age, in the 1960s, reduced the doubling time to about 5 years. And, with the Internet,
the amount of information to which we are exposed currently doubles in less than a year.

Researchers now estimate that humans, with what we currently use of our brain potential, can only
take in 12% of the visual information available. And, for traders and investors the situation is at
an extreme. A trader or investor, looking at every market in the world simultaneously, could
easily have about a million bits of information coming at him or her every second. And since
there are usually some markets open around the world at all times, the information flow does not
stop. Some poor traders actually stay glued to their trading screens, trying to process as much
information as possible for as long as their brain will permit.

The conscious mind has a limited capacity to process about 7 (plus or minus 2) chunks of
information at a time under ideal conditions. A “chunk” of information could be one bit or it
could be thousands of bits (for example, a chunk could be the number 0 or a number like 7,941).
Read the following list of numbers, close the book, and then try to write them all down.

34 39 85 93 21 98 43 56 76 53

You probably couldn’t do it because we can only consciously process 7 (plus or minus 2) chunks
of information at one time. Yet we have millions of bits of information coming at us every
second. And with the current rate of information availability doubling every year, how do we
cope?

1
Part I: Are You Doomed to Failure?

The answer is that we generalize, delete, and distort the information to which we are exposed. We
generalize and delete most of the information. For example, “Oh, I'm not interested in the stock
market.” That one sentence takes about 90% of the information available on the markets,
generalizes it as “stock market information,” and then deletes it from consideration.

Psychologists have taken a lot of these deletions and distortions and grouped them together under
the label “judgmental heuristics.” They are called “judgmental” because they affect our decision
making process. They are called “heuristics” because they allow us to sift through and sort out a
lot of information in a short period of time. Heuristics are shortcuts! We could never make
market decisions without them, but they are also very dangerous to people who are not aware that
they exist. They affect the way we develop trading systems and make investment decisions.

The primary way most people use judgmental heuristics is to preserve the status quo. We
typically trade our beliefs about the market and once we've made up our minds about those beliefs,
we're not likely to change them. And when we play the markets, we assume that we are
considering all of the available information. Instead, we may have already eliminated the most
useful information available by our selective perception.

Interestingly enough, William Eckhardt points out in his chapter of The New Market Wizards that
progress in knowledge results more from efforts to find fault with our theories, rather than prove
them. 1 If his concept is true, then the more we tend to realize our beliefs and assumptions
(especially about the market) and disprove them, the more success we are likely to have making
money in the market.

Thus, what are the beliefs and theories that need to be disproved for us to make progress? These
beliefs represent many of the biases that we must overcome in order to make progress. My
journey as a trading coach and as a modeler has certainly involved a lot of disproving the status
quo.

The secret to success is in understanding how these biases affect you, and then turning yourself
into an effective investor/trader. If you try to project what you learn outside of yourself onto the
market, you will not be able to apply any of the principles taught in this book. Money is made
through the personal application of these principles.

Bias 1: Locus of Control—The Lotto Bias


This particular bias has to do with the need for control—a need we all seem to have—so investors
focus on that area of investing in which they think they have the most control—picking the right
stocks. However, it’s really just a bias.

This bias is particularly evident in the lottery game, Lotto. Almost every government that runs a
lottery offers the game Lotto. And, just in case you are not familiar with it, you buy a card and
you get to pick some numbers—usually seven of them. If the numbers you pick match the
numbers that are randomly drawn, then you win the big multi-million dollar prize. People are

2
The Definitive Guide to Position Sizing Strategies

quite willing to play this game in large numbers because 1) they have the potential to turn a one
dollar investment into a multi-million dollar prize (but it usually is a negative expectancy game),
and 2) they get to pick any numbers they want.

Being allowed to pick any numbers you want in the Lotto game is what makes it appealing. In
fact, there is a whole industry that has sprung out of helping people pick the right numbers. First,
there are actually services that help people pick numbers. They are cheap—only a dollar per
pick—and they basically give everyone a different number. But if they help someone win, they’ll
make a million dollars in the next lottery. Second, there are people who’ll read your astrological
chart and help you pick the right numbers. Third, you can buy software that will analyze previous
numbers that have been picked so you can discern patterns and make better picks. And fourth,
you can even buy software that will randomly generate numbers, just like the machine, so you can
pick one of the randomly generated numbers. On top of that, remember that the lottery usually
announces the store at which the last winning number was sold and when they do, people will
flock to that store to buy tickets for the next lottery.

Does this all sound a little familiar? It should because it is very similar to what happens in the
stock market. People think that winning the stock market game has everything to do with picking
the right stock. About 30% of all books on how to make money in stocks have the word “picking”
in the title. Television shows related to the market frequently bring in fund managers or analysts.
And what does the host ask them? “What stocks are you picking for us today?” They might also
give the track record of the person being interviewed.

Last time Mr. X was on the show he picked XY and it’s up 12%. He also picked
CV, but it’s down 26% and he picked TY and it’s down 18%. What happened, Mr.
X? You didn’t do so well last time.

Notice how the presupposition in all of this is that it’s all about picking the right stocks. And,
obviously, my mistake when I bought my first stock, based on this kind of logic, is that I picked
the wrong stock.

The logic that says that success is all about picking the right stocks is so deep that mutual funds
are always at least 95% invested because they feel they are paid to pick the right stocks and keep
your money working for you. Furthermore, analysts are paid huge six figure salaries and their
only job is to analyze the balance sheets of the companies they research so that they can pick the
right stocks. And, by the way, I have yet to meet an analyst who has managed to become a good
trader through picking the right stocks. Some of them are okay as portfolio managers, but very
few become good traders.

Thus, the average investor, armed with this bias that he can control his success by just picking the
right stocks, finds himself in a world in which picking the right stocks is emphasized by everyone.
So when they lose money, they just assume that they picked the wrong stocks or that someone else
(who was giving them advice) picked the wrong stocks. And what typically happens? The
average investor never learns some of the key factors that are important for success—namely, the
Golden Rules of Trading given earlier.

3
Part I: Are You Doomed to Failure?

What to Do About the Lotto Bias

In this particular case, realize that you also have control over your exits. You can exit at your
predetermined stops and almost guarantee that your losses will be 1R or less. You can use trailing
stops to let your profits run. This will almost guarantee that many of your profits will be greater
than 1R. And if you follow these rules, pretty soon the returns that you generate will be enough to
convince you of the wisdom of the Golden Rules of Trading.

Bias 2: The Need to Be Right


The educational process in most industrial countries came about not to really educate our children,
but to develop good workers for our factories and other businesses. When most people worked in
agriculture, we didn’t need a great educational system—it was just for the chosen few. But now
we need “educated workers” to help with our businesses. Sure, we want these highly skilled
workers to be able to think and come up with new ideas. But we also want them to be good
employees and do what the boss wants them to do. So how do we do that? We do it through our
educational process where children learn that the teacher is always right.

Children go to school for 12 to 16 years and what’s emphasized over and over again is that the
teacher is always right. For example, as a child in school, you have to take tests. You learned that
if you got less than 70% right, you are a failure. And you don’t get an excellent mark, an A,
unless you get 94% correct or better on your test. Perhaps you get 95%. When you showed it to
your dad, he responds, “Why didn’t you get 100?” So your dad wanted you to be right as well.

As a result, we grow up with a passionate need to be right. If you are not right at least 70% of the
time, you are ostracized as a failure. But you want to be right 100% of the time so that your dad
won’t criticize you. As a result, you even criticize yourself first so that you can correct the
problem before your Dad starts to criticize you.

Now, let’s apply that to the stock market or to the futures market or to any other investment you
might make. You want to be right and that to you means making money. Let’s say you buy a
stock for $50 and know enough to set a stop loss—you’ll get out if it drops to $45 per share.

But let’s say it drops to $45 per share. You really want to be right, so if you got out you’d be
wrong, or at least feel as if you were. All sorts of thoughts go off in your head. “It’s just a
temporary setback.” “The analysts are predicting a great increase in the earnings this quarter—I
can’t sell now!” “What if this downturn is just a few traders manipulating the market?” “I think
I’ll hang onto the stock and not sell—at least for a few days.”

So you hang onto the stock and watch it fall even further. It drops to $40. Now you have a 2R
loss. If it was hard to take a 1R loss, it’s even harder to take a 2R loss. And all the same
arguments apply. Thus, you hold onto your stock.

4
The Definitive Guide to Position Sizing Strategies

Now the stock drops to $35 and you have a 3R loss. You know you really should get out, but now
your portfolio is down $4,000. You can only write off $3,000 in losses, so you’d better keep this
stock. You know it will turn around. However, you have a good solution to keep away the
anxiety of watching yourself lose money. You won’t watch it anymore. You’ll look at it in six
months and by that time perhaps you’ll have made a lot of money.

There is an old joke about the man who was dreaming about some “evil” entity that was stalking
him. It kept getting closer, no matter how fast he ran. It got nearer and nearer. Finally, when the
entity was almost on top of him, and he felt sure he was doomed. He turned to plead for his life,
and what did he see? He saw the postman handing him an envelope, saying, “It’s just your
brokerage statement.”

Perhaps, now you can understand why a psychologist and an economist won the Nobel Prize in
economics for basically showing that it was very hard for people to take losses. People, according
to those Nobel winners, become much more “tolerant of risk” when they are behind. Obviously,
people have trouble cutting losses short. But that’s only half the golden rule. The other half is to
let your profits run. The Nobel winners also showed that people tend to tolerate little risk when
they are ahead, making it difficult to let profits run.

So let’s go back to our bias—the need to be right. What happens when you are right about your
investment and it starts to go up? The golden rule says let your profits run—let it go up more.
But you have a strong need to be right. Your $50 stock has gone up to $55 and if you sell now,
you’ll be right and have a profit.

However, you know you should let your profits run and to do that you’ve got a 10% trailing stop.
Now that the stock has reached $55, you won’t sell it unless it drops $5.50 to $49.50—your
trailing stop level. However, suddenly your stock starts to drop. It drops to $54 and then to $53.
You get nervous because your profit is slipping away. Now it drops to $52 and then to $51. You
feel tied up in knots. It’s getting close to your stop and if you get stopped out, you’ll have another
loss. You’ll be wrong. Suddenly, it drops to $50.50, and that’s enough for you. You sell the
stock quickly for a $0.30 profit after costs. You really feel proud of yourself because you made
money.

So what just happened here? Our investor, because of his overwhelming need to be right, sold out
for a minimal profit. The stock actually dropped to $49.90 and then turned around and kept going
until it hit $75. But our investor was happy because at least he didn’t lose any money.

Notice what he’s done here. He’s cut his profit short and let his loss run. And isn’t that exactly
the opposite of the golden rule of trading? What do you think your trading profits would look like
if your results were similar to those shown in Table 5-1?

5
Part I: Are You Doomed to Failure?

Table 5-1: Typical Investor R-multiples


Resulting from the Need to be Right
Trade # R-Multiple
1 +0.1R
2 −3.0R
3 +0.2R
4 +0.2R
5 +0.4R
6 −4.0R
7 +0.2R
8 +0.1R
9 +0.3R
10 −3.0R
Total −8.5R

Notice that because of the bias to be right, our investor has managed to only have three losers. But
those three losers total −10R. Our investor is right 70% of the time with seven winners. However,
those seven winners total +1.5R. And the net result of our investor’s bias to be right is that he is
down −8.5R after ten trades. Thus, if he were investing about 1% in each trade, he’d be down
about 8.5%. Not a very good result for someone who was right 70% of the time—just above the
failure level. And as he wonders what went wrong, he thinks to himself, “Perhaps I picked the
wrong stocks.”

Yet, let’s look at the opposite situation. Let’s say that our investor made money three times out of
the ten trades, two 3R gains and one 4R gain. He lost money seven times out of ten—all 1R losses.
This is shown in Table 5-2.

What’s the net result for this person? Well, they are right 30% of the time, but the net result in
terms of R is +3R. Had they risked 1R one each trade (and about 1% of their equity), they would
have been up about 3% at the end of 10 trades. Now can you begin to see why the need to be right
bias can be so deadly to your bottom line?

6
The Definitive Guide to Position Sizing Strategies

Table 5-2: Typical Investor R-multiples


When Following Golden Rule
Trade # R-Multiple
1 −1R
2 +3R
3 −1R
4 −1R
5 −1R
6 +4R
7 −1R
8 −1R
9 −1R
10 +3R
Total +3R

So now we have both halves of the research done by the Nobel Prize winners in economics. People
tolerate risk more when they are behind (i.e., they won’t cut their losses) and tolerate risk less when
they are ahead (i.e., they won’t let their profits run). And the net result is most people have trouble
making money in the market.

So what can you do about your need to be right? Instead of focusing on being right, focus on
not making any mistakes, where a mistake occurs when you don’t follow your rules. Your rules
should be the golden rules of trading:

• Always know your exit point, the point at which you’ll get out in order to preserve your
capital, before you enter a trade. And if you don’t take such losses when they occur,
consider it a major mistake.
• Always at least keep some sort of trailing stop so that you can let your profits run. And if
you find yourself taking profits too quickly just to make sure you don’t lose money, then
that’s another major mistake.

If you consider breaking these rules as being wrong (i.e., making a mistake), you’ll find that
suddenly you can make money—big money—in the stock market or any other investment field.
And let me repeat the major lesson from the last chapter, because it applies here as well:

In short, you now think in terms of probabilities and statistics. And as a result, you can pay
attention to just following your system, and making as few mistakes as possible, because when
you do that, you “know” what your results will be in the long run.

7
Part I: Are You Doomed to Failure?

Bias 3: Percent Gain


Imagine the headlines….

If you’d taken this recommendation, you’d have been up 150%.

If you had taken all of my recommendations this year, you would have turned
$10,000 into $40,000

XYZ, after I recommended it, went up 300%.

When each statement is made you visualize your entire portfolio being up that much. Instead of
thinking XYZ went up 150%, you think of your portfolio being worth $250,000, instead of
$100,000. However, that would only occur if you invested everything you had in that particular
stock and managed to get the exact amount of profit that was reported. And what’s wrong with
that logic? If you invested everything in that particular stock, your risk would have been huge.
No one should take that kind of risk on a single stock.

Let’s look at what a stock being up 150% really means in terms of an R-multiple.

Say you bought the stock with a 25% trailing stop. You bought it at $10 per share with an initial
stop loss at $7.50. The stock is now up 150%, meaning it is now up to $25 per share. You have a
paper profit of $15, compared with an initial risk of $2.50, which means you are really up 6R in
the stock.

Just because the stock is up 150%, doesn’t mean that you’ve sold it. At $25, your tailing stop is
now at $18.75. Hopefully, it’ll go up more. But if you get stopped out, your total profit shrinks to
$8.75. When you compare this with your initial risk of $2.50, it means you have a 3.5R profit.
And that means if you risked 1% of your equity on the trade, you will make 3.5%. That is a far
cry from thinking that your portfolio has moved from $100,000 to $250,000—but that is what
most people envision when they read this headline, “if you invested in this stock, you’d be up
150%.” You’d probably find that if you invested 100% in each recommendation that you’d blow
out your account very quickly.

Are you beginning to see how this bias works? More importantly, can you see how much better
your thinking would be if you thought of your results in terms of R-multiples or risk-reward
ratios?

Let’s look at the next example: “If you had taken all my recommendations this year, you would
have turned $10,000 into $40,000.” This is another real headline from an advisory service.
However, when I made some inquiries as to what it really meant, this was the answer:

If you had risked $10,000 on every trade recommendation that was made this year, then at the end
of the year you would have been up $40,000. If you now translate that into R-multiples, the
statement becomes “If you had taken a 1R risk on each recommendation during the year, then at

8
The Definitive Guide to Position Sizing Strategies

the end of the year you would have been up by 4R.” Let’s say that this advisor made 20
recommendations. That means the expectancy of his trading recommendations was a paltry 0.2R.

Looking at the original statement, you see your account up 400%. But if you demand enough
information so you can think in terms of R-multiples and expectancy, you discover that it is a poor
system with an expectancy of 0.2R. If you risked 1% on every trade, you’d only be up 4% at the
end of the year.

Now let’s look at the third recommendation, if you’d bought XYZ, it went up 300%. Again, with
this one you see your account up 300%. However, let’s assume that in this case it was an option
trade. Your risk was the entire amount of the option contract. Your eventual profit was 3 times
your initial risk, but since your initial risk was everything, your net profit is a 3R profit. Thus, we
suddenly move from seeing our portfolio up 300% to realizing that we are probably up 3%
because of this one trade.

And, when an adviser tells you about all of the trades that went up 200% or 300%, they are not
telling you about the losses. Thus, you have no idea about the real expectancy of the system or
the real performance of the portfolio.

So let’s say an advisor makes the recommendations in Table 5-3.

Table 5-3: What if Your Newsletter Had the


Following Recommendations?
Trade
Result of Trade R-Multiple
Recommendation
Buy GE at $38 Loss to $28 −2R
Buy IBM at $60 Loss to $50 −2R
Buy GM at $45 Loss to $40 −1R
Buy CREE at $15 Gain to $45 6R
Buy VLO at $75 Loss to $67 −2R
Buy TSRA at $41 Loss to $29 −3R
Buy BHP at $65 Gain to $75 2R
Buy AAPL at $28 Gain to $82 8R
Buy WRF at $33 Loss to $16 −5R
Buy HD at $64 Loss to $58 −2R
Total Gain/Loss −1R

During a period of six months, the advisor’s overall track record is negative 1R. What does he tell
you? In April we bought CREE and sold it two months later for three times what we paid for it.
We also bought Apple and sold it for a nearly a 300% gain. Wouldn’t you like that kind of
success?

9
Part I: Are You Doomed to Failure?

And what’s your reaction to that? “Wow, I could have bought CREE and tripled the value of my
portfolio in two months.” Would you have? If you had bought the entire portfolio, as
recommended, you’d have been down. But the advertising doesn’t say anything that is incorrect.
It just leads you to think that their performance is much better than it really is.

In late 2005, the media announced a merger in which Valero was planning to buy out Premcor and
become the largest refiner in the United States. One advisor had both stocks in his portfolio
earlier in the year. However, both stocks were stopped out of the portfolio a week or two prior to
the merger. However, when the merger was announced, this is what was sent to potential
subscribers.

“One of our stocks recently bought out another one and both stocks had huge jumps in
price on the announcement. You could have made huge profits in both of these stocks
had you followed our recommendations.”

Again, can you see how this bias would hurt most people, especially since he sold both stocks
prior to the merger? However, in this case the solution to the bias is simple. Don’t believe
anything anyone tells you unless they can show you their track record in terms of R-multiples or
as data that you can convert to R-multiples! Otherwise, they are just telling you about a portion of
their recommendations and framing it so you imagine huge gains.

As a result, we recommend that you convert everything that people tell you about their
performance into R-multiples. What was the initial risk? What was the reward-to-risk ratio (i.e.,
R-multiple)? Determine the SQNSM and then see what really happened. And if you do, ask,
“How does that SQNSM compare with other SQNsSM I’ve seen?” For an example of this, see the
2nd Edition of Trade Your Way to Financial Freedom.

Bias 4: Lots of Input Says the Same Thing


This is another significant bias related to the amount of information to which you are exposed.
Typically, the more people are exposed to certain information, the more likely they are to believe
it. Yet, it could be the same information (i.e., from the same source). For example, let’s look at
the idea that “stock picking is important to investment success. Someone develops a story about
how some guru made a fortune picking the right stocks. Let’s say that all of the news wires carry
the story, so you read four different versions of the same story written by four different people.
Now, one source started the story but because you are exposed to it four different times, your
conclusion is “It must be right/true/correct.”

A huge number of sources say that picking stocks is important. For example, I looked up “picking
stocks” in Amazon.com and the inquiry returned 158 items, including

• How to Pick Stocks Like Warren Buffett by Timothy Vick, and


• Pick Stocks Like Warren Buffett by Warren Boroson.

10
The Definitive Guide to Position Sizing Strategies

Notice how there is an assumption here that Warren Buffett, considered by many to be one of the
world’s greatest investors, makes his money by picking the right stocks. However, Warren Buffett
didn’t write the book—someone else did. Using his name and including “picking stocks” in the
title makes it seems as if the key to success is picking stocks. And you don’t have to read the
book to assume that—you just have to look at the title.

Here are a few more:

• Michael Sivy’s Rules for Investing: How to Pick Stocks Like a Pro by Michael Sivy
• How to Pick Stocks by Fred Frailey
• World’s Greatest Stock Picks of All Time by W. Randall Jones
• Investing Smart: How to Pick Winning Stocks with Investor’s Business Daily by
Dhun Sethna
• Pick Winning Stocks by Edward Mrkvicka.

There are many more books with stock picking in the title. These are just to show you the
prevalence of the bias.

However, the topic is even more common on television:

• Wall Street Week always has its panel of experts who pick the stocks they like.
• CNBC has programs like “Stock Picking Friday.” In fact, stock picking is
predominant on CNBC and I have never heard one expert say, “I like this stock, but
I’d sell it if it dropped to this level.”
• Bloomberg will also interview experts and ask them which stocks they like.
• CNNfn (which no longer exists) would frequently interview people to find out
which stocks they were recommending. And my guess is that Rupert Murdock’s
new Fox Business News will strongly feature stock picking.

The list goes on. If you just watch television to determine how to invest, you would be sure that
the key to success was picking the right stock.

What to Do About Lots of Input Saying the Same Thing

Again, if you do what we suggest in this book and have enough confidence in yourself and your
system, it shouldn’t matter what other people say. At tops and bottoms of markets, most people
are always wrong, so do you really want to listen to what most people say?

Bias 5: Authority
We believe people who are in authority. If the analysts say so, they get paid six figure salaries, so
it must be true. I actually pointed it out in the last bias. Two books on picking stocks had to do
with how Warren Buffett picks stocks. In fact, there are nearly a dozen books that have been

11
Part I: Are You Doomed to Failure?

published on Warren Buffett. I even cover some of his style of investing in Trade Your Way to
Financial Freedom. Furthermore, every author makes the assumption that Buffett’s success is
because he is the key stock picker. And if Warren thinks it’s so, we believe it must be so.
However, Warren Buffett has written none of those books and I’m sure that if Buffett told the
truth about how he invests, he’d also emphasize his exit strategy. Now in one sense he doesn’t
have an exit strategy because he buys stocks that are tremendously undervalued. If they meet
those criteria, he will buy them and keep them. But in another way, he does have an exit strategy:
when it becomes clear to him that the stock he’s bought is now overvalued or that the reasons for
his investment have changed, then he’d probably sell it quickly.

People also assume that when analysts and fund managers talk about the importance of stock
picking that these people are authorities. Consequently, it also carries a lot more weight when
these people give an opinion.

What to Do About the Authority Bias

The answer here is obvious. If you do the sort of analysis of your system that we’ve recommend
here, then you don’t need any authority other than your own data. It will give you the answer to
each of the following:

• How to trade? You follow your system because you are confident in the results.
• When do you exit? Your system predetermines that prior to each trade (and even Warren
Buffett does this indirectly by knowing he’ll sell when his company is no longer a good
value).
• How do you pick investments? You don’t pick anything, your system trades when it gets
a signal. Furthermore, you understand that picking stocks and your trading system’s entry
are only a small part of what it takes to be successful.
• How do you predict the future? You can’t predict anything except that you will make
money in the long run. You don’t even know whether your current trade will make
money. In fact it probably won’t because it only makes money 39% of the time.
• What if someone says you are wrong or stupid or crazy doing what you are doing? If
you have confidence in your system and its long-term results, then you won’t care what
other people will say.

Bias 6: Prediction and Understanding


One key need most people have is the need to understand. One of my clients, Joe, claimed that he
had the most difficulty with the market when he got into a position and didn't understand what was
going on. As a result, I asked him a number of questions. “How often are your positions
winners?” His response was that he was right about 60% of the time. “When you don't
understand what's going on, how often do you come out a winner?” This time his response was
that he almost never came out a winner when he didn't understand. I then said, “Since your
system isn't much above chance, you probably don't understand that much about the markets
anyway. But when you clearly are confused, you should just get out.” He agreed it was probably
a good idea.

12
The Definitive Guide to Position Sizing Strategies

When you think about Joe's trading system, however, he really didn't have one. Why? Joe was so
concerned about understanding that he didn't have clearly defined exit signals that told him 1)
when he was wrong so he could get out and 2) when to take his profits.

Most people still need to make up elaborate theories about what is going on in the markets. The
media is always trying to explain the market even though it knows nothing about the market. As I
was working on this section of this chapter, a 91.52 point drop occurred in the Dow. The next day
the newspapers were filled with statements like:

“Investors, spooked by prospects of an economic slowdown, switched en masse


Tuesday to what's become an alluring bond market. The stock market sell-off was
accelerated by computerized program trading.... Money managers are making a
major shift all at the same time, that's why we're seeing such a heavy surge now.
When it fell, it triggered a rush of computerized selling.... Wall Street now believes
that the latest Fed rate increase will slow the economy. That's good news for the
bond market, which hates inflation because it erodes the value of fixed interest
bond payments. But it's bad news for stocks. There's a growing perception that
maybe the rising rates we've had could have an impact on the economy, which
could lead to some corporate disappointments.”

The “need to understand” bias becomes even more elaborate when it comes to trading system
design. People manipulate daily bars in any number of strange ways and then develop even
stranger theories to explain the market based upon those manipulations. The resulting theories
then take on a life of their own, but have little basis in reality. For example, what is the rational
basis for Elliott Wave Theory? Why should the market move in three legs one way and two legs
the other?

When you think about academic theories about the market, those theories are all based upon
predicting the market. Fundamental analysis is devoted to determining the fundamental
characteristics behind the market. Some people believe that when you understand these
fundamentals well enough, you can trade well because you know the factors influencing the
market. In fact, most academicians believe that the markets are totally efficient if you could just
understand the fundamentals. Anything else that might affect the market is just considered to be
random noise.

Some people rebelled against fundamental analysis and developed technical analysis. Technical
analysis amounts to trying to predict the market by looking at pictures of price bars from the
market’s past. Market technicians believe that if you draw enough lines and observe enough
patterns, you will eventually be able to perfectly predict the market.

Now that the Dow 30 has 300-point moves with some regularity—suggesting that the market is
not efficient and random—a new field of study is beginning to replace fundamental analysis. That
new field is called behavioral finance. It attempts to predict changes in the market by studying the
inefficiencies in human decision-making. In other words, psychologists and economists study

13
Part I: Are You Doomed to Failure?

some of the same inefficiencies that I am pointing out to you to determine why the market is so
unpredictable. However, the value in understanding these judgmental heuristics comes from
neutralizing how they impair you. When they no longer impair you, then you have a chance to
make very high return rates with low-drawdowns.

I went to a conference on psychology and the markets in Frankfort, Germany in 1997. Numerous
presenters talked about various ways that human decision making was flawed and how that might
be better used to predict the markets. One even said that what our traders were doing was
impossibleno one could consistently make over 50% in the market. All of the presenters missed
the point. People don’t make money by predicting the markets. They make money by cutting
losses short and letting profits run and by using proper position sizing to accentuate those effects.

The secret to success is in understanding how these biases affect you and in turning yourself into
an effective investor/trader. If you try to project what you learn outside of yourself onto the
market, you will not be able to apply any of these principles we teach in this book. Money is
made through the personal application of these principles.

What to Do About the Prediction and Understanding Bias

Do you really need to understand how markets work? No you don’t. You only need to understand
how the concept that you are trading works. For example, if you are a trend follower, all you need
to understand is that the markets will occasionally move in very large trends and if you can catch
the big moves, you’ll make a lot of money. You have a system that does that, so that’s all you
need to understand about the market.

If you are a value investor, then all you need to understand is why something is undervalued and
be confident in your ability to determine that. The other two things you need to understand are (1)
when your investments are no longer undervalued, meaning it’s probably time to sell, and (2)
when you might be wrong about your evaluation so you can safely abort and preserve your capital.
You don’t need to understand the market at all. Warren Buffett doesn’t—he thinks the markets
are irrational—so why do you need to understand them?

Similarly, no matter how confident you are in your system, you will have trouble making market
predictions. But you don’t have to. You know your R-multiple distribution and you have its
expectancy, standard deviation, and SQNSM. That information will help you determine what to
expect from your system in the long run. And as long as you position size to avoid any worst-case
disasters, you should be able to achieve that expectancy. Do you need to predict anything else?

Are you beginning to see the importance of this kind of thinking and how it can steer you away
from what works? When a pollster predicts how the American population will vote, he doesn’t
necessarily understand why. He just knows what the likely outcome of the vote will be. You have
enough information to know the likely outcome of your system and that’s all you need.

14
The Definitive Guide to Position Sizing Strategies

Bias 7: Wanting Lots of Facts


About 75% of the population have a sensory/detail orientation, while the other 25% have a big
picture orientation. The sensory/detail orientation people have a tremendous bias that keeps them
from trading successfully, which I call the “wanting lots of facts” bias. They want lots of facts and
evidence to support their decisions, whereas the big picture people want to understand how it all
fits together (i.e., the big picture) and then draw their own conclusion. Now how do you think this
affects the two types of people?

Let’s say you went to an investment talk in which some guru was telling you about his Holy Grail
indicator. He might show you something like the indicator shown in Figure 5-1.

His pitch is that he has a magic indicator and when the price goes above that indicator and hits a
new 40 day high (as determined by his software), then look what happens to the price. Now one
chart might not tell you much, but our guru will show you 50 such examples—all followed by a
substantial price increase. And if you are one of the 75% of the population who needs a lot of
facts, then you just got what you needed. Now you know how to pick the right stocks to give you
a lot of money. You’ve been shown 50 examples that his software, with its magic indicator, leads
to higher prices. That’s enough information to convince you it works. You buy the software for
$3,000 and you start to make a lot of money, right? No, quite the opposite is true.

First, you only saw 50 examples in which the price went up. You did not see the examples in
which the price did nothing or went down. As a result, from our guru’s pitch, you have no idea
that the way you make money is through your exits and that the key to meeting your objectives (as
discussed in Part III) is through position sizing. People who need the big picture might pick up on
this, but people who just want lots of facts probably won’t until it’s too late.

What to Do About the Needing Lots of Facts Bias

If you understand the information in this book, but you still need lots of facts in order to be
comfortable, then perhaps trading isn’t for you. Otherwise, simply make enough trades following
your system with low position size until you are convinced that what we are saying is true.

Figure 5-1: The First Piece of Evidence

15
Part I: Are You Doomed to Failure?

Other Biases That Influence Being Right


In the remainder of this chapter, I’d like to focus on the issue of wanting to be right. It’s often
been said that most traders would rather be right than make money. So let’s explore what causes
this to happen.

So now imagine someone who desperately wants to make money in the markets. It’s this person’s
passion. They get a software package that has lots of charting ability and they pour through chart
after chart.

They start looking at the big moves in the market, wondering what those moves have in common.
First, they notice that many big moves follow a consolidation period—not always—but often
enough that it catches their eye. So their first trading idea is to trade moves breaking out of a
consolidation period.

But how do you know what’s a real move or not? Suddenly, it hits them. There is a four-bar
pattern that seems to occur on about 70% of the patterns they see. “That’s it!” they exclaim. And
a new trading idea is born.

Now while this process might be better than what the average person is doing—buying
investments simply because of a news story or a guru recommendation—it still has some major
flaws in it.

Bias 8: The Law of Small Numbers


If you want to find something, such as a pattern that will lead to a big gain, it’s easy to do so. Our
minds naturally gravitate toward finding what we want, in fact creating it, out of chaos. As a
result, we tend to see patterns where none exist, and it only takes a few well-chosen patterns to
convince people that the pattern has meaning. In the example given above, our trader found a
great pattern that he thought would lead to success. In reality, he only found six examples of this
pattern, but that was enough to convince him that the pattern was real and decide that he had a
trading system.

However, here’s what he was missing:

• He only saw the six patterns that work and decided it was real. What he didn’t do
was look at several hundred consolidations to see how often the pattern appeared
and whether it always preceded a new trend. If he can come up with data that said,
“Out of 300 consolidation periods, this pattern appeared in 213 cases prior to a new
upmove,” then he would at least have a reasonable idea that the signal was real.

16
The Definitive Guide to Position Sizing Strategies

• Second, and this is a common bias, he didn’t look for how often the pattern leads to
failure. How often does it occur and not lead to an improvement? Does it occur in
non-consolidating periods? What happens when it does? For example, he might
have developed a computer program to screen his data and found that the pattern
occurred with some regularity, about once every ten days. Thus, in the same time
period that he found 213 patterns leading to up moves, there were actually 7,124
other examples of this pattern that did nothing. Suddenly, we have a pattern that
only works about 3% of the time.

• Now this problem might be fixed by saying, “I’ll screen for a consolidation pattern
first and then look for the four-bar pattern.” This might make it workable. But
there is still the matter of how often the pattern led to up moves. And when you
check this out, it turns out that in the 300 consolidation moves there were 732
examples of the pattern. Thus, while 213 of them lead to up moves, the other 519
examples lead to nothing.

At this point, the pattern doesn’t look so good at all. But even if it did, we’d only have one part of
a trading system—a filter and an entry. A full trading system also needs a worst-case stop, an exit
plan, and good position sizing.

So now you have some idea what the mind can do for you when you want to be right and you
don’t consider all of the issues involved in good trading.

Furthermore, people only see the patterns that lead to success and not the patterns that lead to
failure (i.e., big losses). Imagine what this one bias could do to convince you to buy a stock with a
certain pattern.

Again this particular bias is fixed by following the recommendations you’ve just been given
in the book: learn to adopt a statistical approach to the markets and have a goal of not
making any mistakes, where a mistake means not following your proven rules.

Bias 9: Once We Think We’ve Got It, It’s Hard to Get Rid of It
Once you believe you have found a pattern and become convinced that it works (by means of the
law of small numbers), you will do everything you can to avoid seeing evidence to suggest that it
doesn’t work. For example, once you found the pattern described above, most people would be
very reluctant to see any sort of evidence that says it doesn’t work.

When you read the example above, you probably say to yourself, “Sure, it’s really important to do
all of those things to determine if what I’ve found is meaningful.” But the bias most people have
is to totally avoid doing anything like that. Once you’ve found it, you don’t want to know that
you really haven’t found it.

17
Part I: Are You Doomed to Failure?

There are numerous examples of this:

• If you believe that stock picking is the key to success, you’ll avoid evidence that
suggests it doesn’t work.
• If you think you can make money with options because of the high leverage and
limited risk, you’ll keep trading options despite loss after loss.
• I’ve even seen traders who develop a specific arbitrage strategy that has given them
a real edge. They trade it and make a small fortune and then the strategy stops
working. They’ll even tell me the strategy no longer works, but because of this
bias they keep trading it and lose a lot of money. Perhaps they need real-world
verification that the strategy doesn’t work.

What to Do About the What I Know Is Right Bias

Let me ask you a simple question. Do you believe what I’ve told you about how to evaluate
systems? If you believe that, then that’s all you need to know. If you don’t believe it, then test it
out for yourself. And if you don’t want to do that, then perhaps trading or investing is not for you.

Bias 10: Representation


Is reality what it really seems to be? As someone trained in how the brain works, I can tell you
without knowing about such biases that it is not. Our brain just sees patterns of light that trigger
cells to go off in the brain. We don’t know that something is a book or a ship or a bar chart until
we are trained to recognize it.

So how does this pertain to investing? When people see a pattern in the market, is it really that?
We already saw this example played out with the law of small numbers. But let’s jump further
into what we are actually doing.

When people see something like Figure 5-2, they just assume it represents the market. First, the
description says it is a chart of the S&P 500 so it must represent the market. But does it? When
something is supposed to represent something, people assume that it is that thing.

18
The Definitive Guide to Position Sizing Strategies

Figure 5-2: A Bar Chart

Think about it. That chart collapses months of data into simple bars on a page, but you are willing
to assume that it represents the market. Do you really know what was going on? Who bought and
who sold? Who wanted to buy and who wanted to sell? Or more importantly, what’s going on
right now? We assume that it is somehow in that chart. But that chart isn’t the market. The chart
is just a representation of the stock prices presented in some easy to understand manner. And a lot
of information is deleted in that bar chart.

However, when we start to think that our representations are real (and we all do this), it clearly
distorts our thinking. And most people take it one step further because they do things to the data
(i.e., draw trendlines; determine Fibonacci numbers; determine moving averages), which they
think represent the market even more. But in reality, the more transformations you do on data, the
less likely it is to represent the market.

In reality, the more transformations you do on data, the less likely it is to represent
the market.

You might be feeling that “Van is full of it” by making such statements. “Of course, that
represents the market!” But isn’t that Bias 9 acting in your head?

The only safeguard that I know of for this bias is to step back from everything, be in the “now”,
and just notice what is actually happening. And you can do that if you have confidence in the long
term results of your trading.

I’ve always recommended that your business plan for trading include worst-case contingency
planning. Part of your worst-case contingency planning should center on this particular topic.
What if something I think is real, really isn’t real? What are the implications for my trading?

19
Part I: Are You Doomed to Failure?

Conclusion
Perhaps you can begin to see why I tend to gravitate toward the notion that everything is
psychology. The more you understand this, the more you realize that at some level you are
responsible for everything you experience.

And in my opinion, that’s the first key to being a great trader. You must own your own
performance. You must believe that your system will make money long term because you’ve
taken a valid, reliable sample of your system’s R-multiples. You’ve determined its SQNSM for
each kind of market and you know what to expect in the future. You have specific objectives and
you are going to use position sizing to meet your objectives based upon the methods illustrated in
Part III. As a result of this process, you simply concentrate on the now. Are you doing the
process? Are you following your system or are you making mistakes?

NOTES
1
Schwager, Jack, and William Eckhardt. New Market Wizards. Harper Collins, 1992.

20
The Definitive Guide to Position Sizing Strategies

Part II

Position SizingSM Software Examined

I’m often asked the following question: “What kind of software do you recommend to help me
with position sizing?” I’ll give you two answers to this: a short one and a more lengthy
explanation. The short one is that every software product I know of has some drawbacks. The
long answer you’ll probably understand after I give you some history.

My Experience with Position Sizing Software


The first trading software was either developed to screen stocks (i.e., to help you with the “stock
picking” job that most people believe is important) or to optimize a trading system to fit the data.
The first type of software simply searches the universe of stocks to find some technical or
fundamental criteria that you think are important for stock picking. The second type of software
allows you to overlay all sorts of technical indicators on a multi-year set of trading data to develop
some sort of system that will give you a great profit on the instrument you are testing.

The problem with both of these types of software is that they don’t operate the way you do as a
trader using position sizing. With position sizing you must make decisions with a portfolio of
positions time period by time period. That’s not really compatible with how other types of trading
software (that most people want) operate. Thus, there is a basic incompatibility between position
sizing software and trading software.

The first person to seriously address this issue was Bob Spear who developed a product called
Trading Recipes. Trading Recipes worked on a set of data as a whole, so Bob didn’t really solve
the incompatibility problems I just mentioned. He did, however, have a position sizing overlay to
work with the trading system you’d developed. Nevertheless, because the money management
was basically an overlay to traditional “one trade at a time” type software, it was impossible to do
anything that was time dependent in position sizing such as scaling in and out. Bob and I worked
together for a while and I discovered the seriousness of the incompatibility problem. Trading
Recipes was a DOS-based product and he found that the job of converting it to Windows and
making it do “real” position sizing was overwhelming for him in the mid-1990s. However, his
new Windows version, Mechanica, is now available.

My next adventure into position sizing software was when a software developer from England
developed a product called Athena that would do everything that was in the original Special
Report on Money Management. The software was great, but very expensive ($12,500). It
basically linked with Trade Station® to combine systems with multiple position sizing models.
Moreover, it had basic trading models built into it, in addition to “thinking” in terms of R-
multiples. Athena had a channel breakout system built into it and also a random entry system.
Many of the models from the original Money Management Report were tested with that software.

21
Part II: Position Sizing Software Examined

Athena also had major problems that prevented it from ever becoming a viable product. One of the
problems was the lack of technical support for the product. In addition, through my research I also
discovered that position sizing software like Athena could optimize position sizing to do
extremely well with past data and not perform that well in real trading. As a result, I became
much more interested in simulators to look at position sizing.

The first simulator we used was an Excel simulator developed by Frank Gallucci. At one time we
offered a position sizing workshop with Excel products that Frank had developed. We stopped
doing these workshops, which included the software for free, simply because there was not
enough demand for them.

And finally, Chris Anderson developed a much more sophisticated simulator called Know Your
System. I used this software to do my position sizing research, and have saved you a lot of time by
giving you the results in this book. However, it is not available for sale for two primary reasons.
First, there are major assumptions made with R-multiple simulations that could be violated by real
trading. People could make some financially ruinous conclusions if they didn’t understand those
assumptions. Second, my company is not a software company and my staff is not capable of
doing any sort of technical support for the software.

I originally wrote the Definitive Guide to Position Sizing with the idea of bundling it around Know
Your System. When we made the decision not to move in that direction, much of this book had to
be rewritten. Instead, I have relied on using the System Quality NumberSM to help you with
guidelines for what you can do with position sizing. We are also putting out the Secrets of the
MastersTM Trading Game version 4.0, which has a lot more simulation capabilities, including
allowing you to see your drawdowns in terms of R-multiples.

Overall, position sizing software has many problems, but the same can be said for trading
software. My opinion, prior to writing this chapter, was that you probably need to develop your
own software, or learn to program in Excel to get what you need in terms of position sizing. In
fact, almost every really good software solution seems to require that you learn to do some
programming. You probably didn’t want to hear that, but that’s just the way it is.

Nevertheless, I’ve asked various people to fill out a brief questionnaire about the software they are
using. In some cases, I’ve asked the developer to fill out the questionnaire. Please understand that
just because the software is mentioned here does not mean that I like it or recommend it. In fact,
at the time I wrote this review, I had not personally tried any of the products mentioned here with
the exception of XLQ.

As a result of doing this review I now believe that there are a number of packages out there that
have become quite sophisticated. And whether you want to find a simple system that works and
allows you to do position sizing or develop an almost custom solution to your needs as a trading
business, there is some software that could meet your needs. Finding the right software is a lot
like finding the right position sizing algorithm. You need to find out who you are, what you want
to accomplish, and then look for the software package that comes closest to meeting your needs.

22
The Definitive Guide to Position Sizing Strategies

Software of this nature usually is “out of date” very quickly. Thus, if I mention that some
software has certain shortcomings or lacks certain features, it doesn’t mean that will be the case
when you read this. As a result, I’ve also included a web site for each product mentioned. I’d
recommend that you go there, read about it, view the software demo (if they have one) or
download the software users manual (if it’s available), ask questions related to what you want, and
then make more informed decisions. In addition, I’ve only included software that is available
commercially and has some support for the end-user.

First, it is important that you understand that there are six different categories of software that
might be useful to you. I have not included software that primarily does screening or system
optimization in any of these reviews. I’ve chosen to include six categories:

1. Software to keep track of trades and help you with expectancy and R-multiples.
2. Simulation software.
3. Position sizing software.
4. System-specific software with some position sizing capability.
5. Multipurpose software with position sizing capability.
6. Advanced software that might save you from having to spend hundreds of thousands of
dollars on custom programming to run your trading business.

We will be reviewing the software that our clients have mentioned under each of these categories.
In each case, when I mention some software, I’ve also put the name of the person who filled out
the questionnaire, gave me enough information so that I could write the review, and, in one case,
actually wrote the review that is included in this chapter.

Software to Keep Track of Your Trades


With software that keeps track of your trades, there are several ways to go. The first way involves
using spreadsheets to do almost everything. Most people who travel this route use Excel, although
one person said he uses Lotus 1-2-3.

EXCEL
If you have Microsoft Office on your computer, then you have Excel on your computer. In this
guide I’ve already shown you examples of Excel keeping track of your R-multiples for each trade.
You can then use the many Excel functions to determine the mean and standard deviations of your
R-multiples as well as your System Quality NumbersSM. And if you use Excel, you can basically
keep a running total of all of this information. For example, look at Table 2-6, which is an Excel
table.

Excel also has the ability to get data from financial websites by simply clicking on the data tab,
clicking on “import external data” and then on “new web query.” This will bring up a box in
which you can enter the address of the web site you wish to go to (for example, Yahoo! Finance is

23
Part II: Position Sizing Software Examined

full of historical data). And once you get to that web site, you can simply import the data you
want into Excel. This requires a certain amount of knowledge of how to use the spreadsheet.

XLQ
If you like the Excel route and don’t mind doing your own programming, then I highly
recommend that you subscribe to XLQ 1. This is basically a whole series of enhancements that
you can add on to Excel that will give you lots of financial formulas and indicators. For example,
there are formulas built-in to do many of the most common trading indicators, such as the Average
True Range, various moving averages, MACD, DMI+ and DMI−, etc. There are over 250
different formulas, including many fundamental values that are added to Excel when you use
XLQ.

XLQ costs $74 ($119 for the enhanced version) with a reduced yearly renewal price and it is well
worth it if you are very competent with Excel. In fact, Ken Long uses XLQ to write and send out
huge reports on ETFs and Mutual funds to his database every evening (www.tortoisecapital.com).
If he did this by hand, it would take 6-8 hours to prepare, but he does it all with XLQ and it only
takes a few minutes of his time each day to run the software and generate the reports.

I would actually follow this same route, but it requires that you really become very competent
working with Excel, and learn how to program macros, etc. That is not one of my skills at this
moment, so I’m not using it as much as I’d like. But if you are considering software that still
requires you to do a lot of programming and you don’t know how to program, then learning how
to program in Excel and how to use XLQ might be the way for you to go. There really isn’t any
training for how to use XLQ except for a demo spreadsheet; however, a fully functional version of
XLQ can be downloaded and used for free for 45 days before the purchase. Ken says that there is
a very active Yahoo users group for XLQ where you can ask questions and get help.

XLQ also has a COM interface allowing you to use all the formulas and data via other programs
with a COM interface or any of the popular programming languages, including Visual Basic, C#,
Access, C++, perl, etc.

For information about XLQ, go to http://www.qmatix.com. In addition, the developer, Leo van
Rijswijk also offers custom solutions.

Stator® Financial Management Software


I’ve included Stator®2 in this review because several of my clients have recommended it as a
strong financial software package. Here’s what one reviewer said:

“I think the strong points of the software don’t lie in its position sizing capabilities (it doesn’t
feature any testing, for example), but in the extensive ways you can present your past trades via
statistical measures and graphical representations. It has all of this plus the ability to handle

24
The Definitive Guide to Position Sizing Strategies

multiple systems and generate a trading diary. Thus, it is a great package for monitoring your
trading.” —Thorsten Reiss

The software is easy to use and will analyze your entire portfolio. It does have percent risk, along
with portfolio heat and group heat, but it really isn’t designed for position sizing. It’s portfolio
analysis software. Here’s what the web site says:

“Stator® provides you with all the tools you need to monitor and analyze your trading performance
so that you can accomplish what all traders strive to do:

“1. Limit losses by practicing sound risk management techniques.


2. Improve weak spots in your trading methodologies.
3. Know your exact Profit/Loss situation at any point in time.
4. Learn from your mistakes so that you never repeat them.
5. Have total control and confidence in your trading systems. [Note from Dr. Tharp: I tend to
doubt the control part.]
6. Find the perfect trading formula suited to your trading style.
7. Find and exploit new trading opportunities from all over the globe.” 3

In addition to all of these benefits, you will also cut down on the amount of time you spend on
simple administration tasks so that you can concentrate on finding more profitable trading
opportunities.

Proper performance management is where successful trading evolves from. For the cost of less
than one single trade you can make an investment which will influence all of your future
trades for the positive.

It has the following important features:

• The ability to work with trading pools (i.e., multiple systems)


• Stop and target management of ongoing trades.
• A tax module.
• Extensive charts and system statistics. They’ve even included the System
Quality NumberSM based upon one of my answers to a question that was
reproduced in Tharp’s Thoughts.
• It allows you to create a trading diary.
• And it does numerous reports.

I have not seen this software myself, but you can visit www.stator-afm.com to learn more. The
website contains over 20 free video tutorials, which should give you a good feel for the software.
The software comes in three editions, ranging in price from $55AU to $495AU (~$417US) and,
according to the reviewer, is easy to use. Also when you purchase the software, you get a number
of bonuses. I was interested to see that one of those was a copy of an article I wrote with Hank
Pruden on the Tasks of Trading.

25
Part II: Position Sizing Software Examined

StockTickr
StockTickr 4 provides an online trading journal, shown in Figure 17-1, that tracks the performance
of your trading system using the R/expectancy model. Trades can be manually entered into the
journal or automatically entered using a simple Application Programming Interface. There are
plug-ins for various software vendors and brokers.

When you enter a trade for a particular stock, StockTickr displays default values that correspond
to your trading history and preferences. These defaults can be adjusted to fit your style. The
values can be easily changed to reflect an actual trade. Changing the values for open price, shares,
stop price, and portfolio value automatically adjusts the percent risked fields so you can see what
your risk would be under different scenarios.

Figure 17-1: StockTickr Journal Entry

Once the trade is entered, it goes into your trading journal that can be accessed from any web
browser with your login and password. StockTickr provides an "R table" that can be accessed by
hovering over the dollar icon, shown in Figure 17-2. This displays price levels that would need to
be reached to meet certain R-multiples for the trade.

26
The Definitive Guide to Position Sizing Strategies

Figure 17-2: StockTickr R-Multiple Table

For each trade, StockTickr automatically generates charts in various timeframes with the entry
point, initial stop, and exit points plotted on the chart. This helps traders determine if they are
moving their stop too soon or not quickly enough. You can also query for charts with various
characteristics, such as 15 minute charts for trades that resulted in a 3R or better gain. This is
shown in Figure 17-3.

27
Part II: Position Sizing Software Examined

Figure 17-3: StockTickr Candlestick Chart

StockTickr also provides a calendar view of your trading that is color-coded, showing more
extreme gains and losses in darker colors. You can write comments for each day and you can
click on the number link on each day to view the trades that occurred on that particular day. This
is shown in Figure 17-4.

28
The Definitive Guide to Position Sizing Strategies

Figure 17-4: StockTickr Calendar

Figure 17-5 is an overview of the expectancy of your trading system per month.

Figure 17-5: Monthly Performance

One of the neatest features of StockTickr is the ability to assign "tags" or categories to each trade.
For example, you might want to track the performance of different strategies such as trading

29
Part II: Position Sizing Software Examined

against the 15 minute bars versus the 30 minute bars. You might assign the tags “long,” “15
Minute Bars” for a long trade you took using the 15 minute intraday bars and “short,” “30 Minute
Bars” for a short trade using the 30 minute intraday bars. This would allow you to track the
performance of your trading system for trades you assigned “15 Minute Bars” versus trades you
assigned “30 Minute Bars.”

You can assign multiple tags to each trade and then access reports based on each tag. Figure 17-6
is a sample report based on the tag that was assigned to trades.

Figure 17-6: Example of Tags

There are a variety of reports available to slice and dice your trading system and figure out what
works and doesn't work, such as the one in Figure 17-6.

StockTickr also allows you to detect trends, and trade based on probabilities of what works and
doesn't work with your trading system. There are new reports being added quite frequently that
give the trader more information about and more confidence in their trading system.

You can find out more information about StockTickr at http://www.stocktickr.com.


I also believe you can do similar things at StockCharts.com, but we did not receive any reviews
for that web site.

Simulation Software

Secrets of the MastersTM Trading Game


The Secrets of the MastersTM Trading Game is the only software product that my company puts
out. The game’s purpose is to help you understand (at an experiential level) the immense impact
that position sizing has on your bottom line. You get to play three levels for free when you
download it from www.iitm.com and then if you activate it, you get to play the remaining levels.

30
The Definitive Guide to Position Sizing Strategies

When people play it over and over again, the typical response is “I learned so much.” But if your
approach is to “try to figure it out,” then you’ll probably discover that there is no answer, simply
because position sizing is as much an art form as it is a science.

You can probably do any sort of position sizing model in this book in the game, but none of these
are built-in because that would defeat the purpose of the game, which is to experience different
methods of position sizing. Instead, to follow some model, you’d have to calculate exactly how to
do it outside of the program and then enter it into the program for each trade as it comes up. You
could even take the last 10 R-multiples as a measure of volatility and use volatility position sizing.
It’s up to you and your imagination.

Yes, we could make it automatic and incorporate many models in there. We could make it do
everything quickly based upon a particular model. But that’s not the purpose of the software.
This software is designed to give you a feel for the impact that position sizing will have on your
trading, one trade at a time. And to do that you have to experience it one trade at a time. This
means calculating your position sizing, entering the trade, and then seeing the results.

One nice feature of the software is that you can plug in the R-multiple distribution from your own
system and simulate that one trade at a time. The Secrets of the MastersTM Trading Game version
4.0 will also keep track of your peak drawdown in terms of R for you because that feature is
required in some of the position sizing models.

Version 4.0 of the game is designed to work well with Windows Vista. We also have a new
realistic feature. In the old edition, there was no slippage, no commissions, no taxes, and no
psychological errors except for some built-in losses. All of these things limit the growth of your
account, but they were not built into the game. People would think the game was unrealistic
because they could make trillions of dollars. The newest version will have a box you can check to
add in most of these obstacles to wealth building for those of you who want more realistic results.

Incidentally, we get people who ask me how to do specific position sizing models with the game
or try to find out how to get through a particular level. Remember the purpose of the game is to
experiment with position sizing so you can get a feel for it. It’s not a problem to solve. It’s
supposed to be a learning experience and you only get that experience by playing it a lot and
experimenting. Asking me what to do (or to comment on what you are doing) in order to get
through a specific level defeats the purpose of the game.

TradeSim®
TradeSim®5 was developed by an Australian company, CompuVision, as an add-on product for
MetaStock. The author said that he developed the software after reading about Athena in Trade
Your Way to Financial Freedom.

TradeSim®, however, is much more than a simple add-on because it allows you to do portfolio
analysis, position sizing, and Monte Carlo simulations. TradeSim®, according to its web site, is a
true portfolio trading simulator and backtester, which analyzes the trades in the proper

31
Part II: Position Sizing Software Examined

chronological order and sequencing, thus mimicking the way that real trades would be executed.
The web site says that it can do dynamic money management (i.e., position sizing) and risk control
at the portfolio level.

So let’s take a look at what one of the reviewers said about the various features. First, it comes
with a basic system to simply demonstrate its functionality, but the software requires either
Metastock or Bull Charts to generate the system signals. My understanding is that you can also
manually input data to do simulations. It’s easy to use, according to the reviewer, but does require
some programming.

In terms of position sizing, it does equal units (i.e., 10% per unit), fixed dollar risk per position,
percent risk, percent volatility, and portfolio heat. It apparently doesn’t do things in which the
position sizing depends upon what happens in the portfolio without some tweaking. For example,
scaling in to positions could be done if you assumed two systems were operating together to
produce the signals. Scaling out is also possible by assuming multiple systems. However, it does
not do any form of market’s money, nor does it do fixed ratio position sizing.

TradeSim® will calculate expectancy and it will provide the data to generate the standard deviation
of R, but you would have to calculate it outside of TradeSim®. It will not calculate the worst-case
drawdown in terms of R.

Apparently, TradeSim® is pretty good with simulations. Based upon comments from the reviewer
and the developer, I think TradeSim® may do simulations with R-multiples, on R-multiples with
position sizing, and on equity curves. I’m curious if it can do that, why it can’t generate the worst-
case drawdown in terms of R.

The software will allow you to run up to 20,000 simulations of the portfolio to generate the risk of
ruin, frequency distributions and standard deviations of key portfolio statistics such as net profit,
percent wins, percent losses, average drawdown, and maximum drawdown. The Monte Carlo
output also produces charts so that you can see the relationship between various system statistics,
such as profit to drawdown.

The Enterprise edition also generates open equity curves so that you can review individual trades
on the chart. You can also review a whole portfolio of systems at one time. And lastly, you can
model slippage with different types of buy and sell orders.

The documentation with TradeSim® is excellent and there is a strong user community with forums
for support. To buy it, you simply pay for it online at
http://www.compuvision.com.au/TradeSim.htm and download it. There are actually three
versions of TradeSim®, the standard edition ($159US), the professional edition ($385US), and the
enterprise edition ($1,199US). What was reviewed for me was a pre-release version of the newest
Enterprise edition (V5.2.0) of the software. You can buy lesser versions, such as a trial to the
standard version and then upgrade to other versions if you think it meets your needs.

32
The Definitive Guide to Position Sizing Strategies

Position Sizing Software

Market System Analyzer


Market System Analyzer 6 is perhaps the purest form of position sizing software that I’ve seen on
the market. Again, I have not used it, so this review is simply based upon users’ comments and
my observations looking at the software web site.

In terms of position sizing, Market System Analyzer helps you determine what your position
sizing should be with a large number of models, including the following:

• Fixed number of shares/contracts


• Units per so much equity
• Percent Risk
• Fixed ratio and a version of that in which you can adjust the speed at which
position sizing increases (GRPS). It probably doesn’t include FRPS the way
we recommend you do it in this book.
• Margin target
• Leverage target
• Equity curve crossovers

It also includes a number of methods we don’t recommend including optimal f, Kelly Criterion,
and a Larry Williams method based upon your drawdown.

The software seems to be missing the ability to do time dependent analysis such as market’s
money, scaling in and scaling out. However, the manual explains that to do scaling in or scaling
out, you simply have to treat each scale in or scale out as a separate trade and enter it into the
software individually.

The software also does Monte Carlo simulations to give you confidence levels on your return rate,
drawdown, return/drawdown ratio, and a modified Sharpe ratio. It does Monte Carlo simulations
on your equity curve (as I understand it), but not on your R-multiples.

It helps you optimize your position sizing to meet various objectives including 1) maximum net
profits, 2) maximum rate of return, 3) maximum average trade in a currency, 4) maximum average
trade percent, 5) maximum profit factor, 6) maximum return to drawdown ratio, 7) modified
Sharpe Ratio, and 8) limiting your maximum drawdown to a percentage of equity.

It also allows you to include trade dependency studies, parameter studies (a graphic position sizing
sensitivity analysis), statistical studies, and has the ability to create trades from statistics. I have
not really worked with any of these studies, so I don’t have a feel for how useful they are.
However, the dependency analysis could be worth the price of the software by itself.

33
Part II: Position Sizing Software Examined

And lastly, you can either import data from TradeStation® (which is not necessarily a plus for me)
and MT Predictor (reviewed below) or enter it as a spreadsheet. One drawback, in my opinion, is
that it takes the data as the total profit/loss on the trade rather than as R-multiples. However, you
can also input your risk, so there might be some way to use R-multiples.

The third version of the software is about to come out and the web site says that it will do portfolio
analysis. The software is priced at $199 right now, so the price is certainly low enough for you to
experiment with if you are so inclined. And you can download a free trial of the software, so
that’s even better. There is also a complete manual online for how to use the software.

Although I haven’t used the software, I’ve looked at the web site (http://adaptrade.com). You
might find it quite interesting and worth your while.

System Specific Software with Position Sizing Capabilities

MTPredictorTM
MTPredictorTM is the software that more people commented upon than any other. There is a
reason for that. When I asked for a review, the folks at MTPredictorTM sent out an email asking
their customers to send me a review. Eventually I just settled on having the people who developed
and sell the software fill out the questionnaire.

This software is a little bit different from any of the others. It essentially offers trend following
resumption methods based on a special Isolation ApproachTM to Elliott Wave and has risk/reward
assessment and position sizing abilities attached to it. When I asked, “What’s an isolation
approach to Elliott Wave?” I was told that it involves a trademarked process that isolates the
simple Elliott Wave ABC correction-to-trend and uses it to enter trades with a small, controlled
risk and high potential reward. This approach has the added advantage of not having to fit the
pattern into a larger pattern or to fit a smaller pattern into the isolated correction. In other words,
it’s a trend-following with retracement type method.

Thus, I thought I’d do a review on both the methods and the software.

The System: There are 5 main types of trade setups (TS) automatically identified by the software
and the developers stress MTPredictorTM is a method, not a system.

• TS1 involves a trend resumption into Elliott Wave 3.


• TS2 involves a trend resumption into Elliott Wave 5.
• TS3 involves a trend resumption into an unspecified Elliott Wave.
• TS4 involves a non-specific trend resumption.
• DP involves a swing trade confirmed by divergence.

34
The Definitive Guide to Position Sizing Strategies

The MTPredictorTM process enables the trader to find a trade, assess its risk/reward outlook,
position size and manage the trade. The standard exit strategy uses automatically-generated Elliott
Wave targets for profit-taking and there is also the option to use their Average True Range (ATR)
volatility stop, adapted from J. Welles Wilder’s work.

With that information in mind, the developers supplied me with 838 trades published daily for
customers between July 26, 2004 and July 22, 2005. These were all day trades lasting a few
minutes to hours. The starting account size was $50,000 (though a minimum of $10,000 is advised
as acceptable). Included were the key TS1, TS2 and TS3 set-ups, on the US index futures and
ETFs and with a minimum +2x risk/reward outlook. Profits were taken according to the standard
exit strategy mentioned above. The data did not include slippage and commissions. In addition,
the R-multiples I was given were rounded to the nearest 0.25R.

The 838 trades had an expectancy of 0.46R with a standard deviation of 2.42R. They produce a
System Quality NumberSM of 5.52, which is excellent and highly significant. However, the huge
number is partially due to the fact that they gave me 838 trades. If I just look at the ratio of the
expectancy to the standard deviation and then base the System Quality NumberSM on 100 trades, it
comes out to 1.91, which is still an acceptable system that makes money at better than a chance
level. A number of our customers said they were using this software (and thus these methods) and
they were quite happy with it.

The Position Sizing Capabilities: According to the developers, the software does percent risk
position sizing, using an integral position sizing calculator in both the real-time and end-of-day
versions. Position sizing is supported in stocks, forex and futures. Percent volatility and group or
portfolio heat are not supported. Scaling in position sizing (pyramiding) and scaling out
techniques are routinely explained to customers, but don’t seem to be a part of the software.

As it is not a standard system, it doesn’t do simulations, but the developers say that they support
and use Adaptrade’s Market System Analyzer, which was reviewed previously. It also doesn’t do
expectancy calculations, but the in-house records (i.e., the ones I reported on) are available and
customers’ own records are often posted to their discussion forum.

Thus, if you want to use a complete process from identifying the trade, through risk/reward
assessment and position sizing to logical trade management, in a relatively automated package,
then MTPredictorTM might be for you. It works on worldwide liquid markets and is supported by
daily training reports focusing on risk control and position sizing. There are also two versions of
the software: the end of day version is $1,995 and the real-time version (i.e., for day trading) is
$2,495. For more information, please go to http://www.mtpredictor.com.

The software requires a moderate level of skill to use, but doesn’t require programming skills,
which may be a plus for some of you.

35
Part II: Position Sizing Software Examined

Multipurpose Software that Includes Position Sizing

AmiBroker
AmiBroker 7 is a fully programmable open-ended system that can probably do most of the things
you want with sufficient programming. But that's like saying SAS or C or FORTRAN, for
example, are capable of doing what you want, too. AmiBroker's advantage over general-purpose
languages is that it has dozens and dozens of precoded features for technical analysis. My reviewer
said, “Someone without strong programming skills who hoped to start AmiBroker and do all of
the things you suggest in the second edition of Trade Your Way to Financial Freedom will be very
disappointed. For example, I have not been able to program it to do simulations with my R-
distributions to help me select a position sizing algorithm to achieve my objectives.”

I looked at the Table of Contents of the Users Manual and the newest version has a position sizing
variable built into it. It gave the following comments about using it.

“For example,

“Position Size = 1,000/ means invest $1,000 in every trade


Position Size = −20 /means invest 20% in every trade (minus means invest a percentage of equity)
Position Size = −100 + RSI() means that the amount invested will depend upon the value of the
RSI indicator with lower value resulting in a bigger investment.”

This is a classic example of how software developers will invent new position sizing models. This
one certainly was not covered in this book, but it also doesn’t make any sense to me, unless it was
somehow used in conjunction with a trend following retracement. But even then, does that mean
invest 100% when the RSI is zero?

There is also a section that says “allow position sizing shrinking, which allows you to still invest if
your available cash is less than the position size algorithm requires.”

I also thought it funny when I read, “Below is an example of the Tharp ATR based position sizing
technique coded in AFL.” And what followed was an example of using 1% risk, where risk was
defined by a trailing ATR stop. Thus, the example was both risk based and volatility based. So
obviously you can do percent risk and percent volatility position sizing.

I also read a section of the Users’ Manual on portfolio backtesting. It gives an example of being
100% invested and dividing your portfolio into X number of equal positions based upon the dollar
amount.

Another section of the Users’ Manual focuses on pyramiding, so obviously you can scale in and
scale out. However, to give you an idea of the focus of the software, the manual gives the
following as examples:

36
The Definitive Guide to Position Sizing Strategies

“1) Dollar fixed cost averaging, so that each month you might buy $X worth of some security.
2) Increasing the position when the profit is greater than 5% and decreasing the position when the
position has a loss greater than 5%.
3) Partial scaling out at profit targets.” 8

Approximately nine pages of the nearly 800 page Users’ Manual were devoted to position sizing
and portfolio testing. I saw nothing in the manual about simulations or R-multiples. However,
that doesn’t mean that someone hasn’t written something for this package that you might be able
to use.

The price is right for this software. It’s a one time fee of $299 for the professional edition and a
$149 fee for the standard edition. The fee includes four upgrades. Also, Amibroker’s language,
AFL, is an open architecture language and my understanding is that there is a strong user group.
The software seems to be compatible with multiple data feeds.

My opinion, after reading the examples in the manual, is that learning the language will involve a
steep learning curve, and this software would not be very useful at all if you did not master the
programming skills necessary to use it. For more information (and to purchase or download a free
trial), go to http://www.amibroker.com.

Since this review was written, I was sent a book entitled, Quantitative Trading Systems 9 by
Howard B. Bandy. The book is written by someone with an extensive math and statistics
background and it gives lots of programming examples using AmiBroker’s AFL language. Thus,
if you plan to use this software, the book is a must.

OmniTrader Professional
I included OmniTrader 10 in this review because of the recommendation of a long time client.
However, when I went to the web site, www.omnitrader.com and took the tutorial, I would not
have guessed it to be a product that does position sizing or simulation extensively.

Anyway, I got the following information from the web site about OmniTrader Professional.

“OmniTrader is the only software that generates buy and sell signals based on the ‘personality’ of
each security in a given list. We call this powerful technique the Adaptive Reasoning Model
(ARM). In just seconds, the program will test all of its 120 built-in methods on each symbol in a
given list to find the precise techniques that are working well. Then, it uses those methods to
generate signals. ARM was invented at Nirvana Systems in 1994, and here is how it works:

“1. A list of symbols is provided to the software. This list can be as short as the S&P 100 or as
large as the entire stock market, depending on how many candidates you want the program to
generate for you.

“2. Press ONE BUTTON in OmniTrader, and it will begin its proprietary analysis. In a matter of
seconds, OmniTrader will test each of its built-in trading methods on each security, choosing the

37
Part II: Position Sizing Software Examined

BEST techniques to use. Using this approach, OmniTrader is able to determine the personality of
each individual security, according to our proprietary Personality of Markets Theory.

“3. The software automatically finds Buy and Sell Signals on each symbol in the list, using the
methods that have been found to work well for each individual security. The result is a set of Buy
and Sell Signals—automatically. No other software offers this automation benefit.” 11

I have no information about the R-multiples generated by this system, and thus can offer no
opinion about the system quality. You are on your own here to do your own research.

OmniTrader also has a simulation mode, but it doesn’t appear to be a Monte Carlo simulator.
Instead, it is a way to practice trading the system to see what would happen before real money is at
stake.

The latest edition of OmniTrader (2007) has a portfolio simulator. And according to Ed Downs
from Nirvana Systems, it does all of the following position sizing methods: fixed size, fixed price,
fixed percent, price to equity, size to equity, optimal f, Kelly, percent risk and portfolio heat. It
doesn’t do market’s money, or scaling in or scaling out.

The portfolio simulator provides full reporting of statistics (e.g., worst drawdown, best trade,
worst trade). The user can simulate all methods at once to pick the best one. And the simulator
generates equity curves that you can look at to determine how smooth they are.

There are several versions of the software 1) a Stock Version for $495, 2) a Futures Version for
$695, 3) a real-time version for $895 and 4) a professional version for $1,948.

Trading BloxTM
Trading BloxTM12 has a number of fairly good systems built into it, plus built-in position sizing. In
addition, it also works with portfolios. Let’s look at the systems first. The built-in systems in
Trading BloxTM depend upon the version you purchase. The “Turtle” version includes three
systems:

• The Turtle System,


• Triple Moving Average, and
• Donchian—a simplified Turtle-like system that does not pyramid.

The Professional and Builder versions add eight more systems and some “trading blocks” that are
not included in the standard version. The additional systems include:

• ADX,
• +DI/−DI
• ATR Channel Breakout—a volatility channel breakout system
• Bollinger Breakout
• Bollinger Counter Trend

38
The Definitive Guide to Position Sizing Strategies

• Dual Moving Average


• MACD
• RSI
• Stochastics

The extra blocks include a strength filter, a MACD filter, a group risk manager, having a profit
target, pyramiding, having a gap open against, chandelier exits, percent risk, and percent volatility
position sizing. If you want to have the flexibility to use all the basic position sizing algorithms,
Trading BloxTM Professional is the minimum version to purchase. In addition to the supplied
BloxTM, Trading BloxTM Professional will allow you to add other BloxTM written by other users or
posted to their support forums.

Trading BloxTM Builder has everything that the professional version has, plus the ability to build
your own new trading blocks. This does require some programming skill in the BloxTM
programming language.

The software allows you to look at R-multiples and do an expectancy calculation. It also has some
Monte Carlo simulation features including the ability to look at what your possible drawdowns
might be as a percent of your equity.

Let’s look at the basic building blocks of this software.

First, the portfolio manager allows you to track securities, commodity futures, and forex futures.
It will also let you dynamically select which markets or stocks are available for trading based on a
portfolio selection algorithm defined in code. And it has the ability to enter the exact
specifications of the futures contracts so that the data can be properly analyzed.

Second, the Entry block allows you to create an order for actual trading.

Third, there is a Money Manager block, which implements the position sizing algorithm for a
system.

Fourth, an optional Risk Manager helps you set risk parameters for dynamically controlling the
risk of your positions. This includes the ability to insert stops, and adjust the position size of open
trades based upon risk criteria that you set. This block allows you to both scale in and out of
positions based on risk criteria.

Lastly, the Exit block allows you to specify your exit conditions. It’s important to know that this
block will accept multiple criteria.

The System Editor, shown in Figure 17-7, is used to create new systems and modify old ones. It
requires no programming skill to do this. All you do is simply drag and drop available blocks into
the system and several entry and exit blocks can be used with each system. Each block can also be
used in multiple systems, so all systems can share a common position sizing algorithm, for
example.

39
Part II: Position Sizing Software Examined

The figure illustrates the concept of building a system by dragging various “BloxTM” into
appropriate spaces. Notice at the top you have a list of available systems, a list of available
trading blocks (or BloxTM), and a place to drag all the components, including the portfolio
manager, the entry signals, the money manager the exit signals, and the risk manager.

The system tester allows you to test one system (or multiple systems) as a portfolio. And if you
use multiple systems, each can be weighted by certain settings. Also, variables like the interest
rate, the initial stop, and slippage can be defined globally so they are valid for each system. In
addition, separate position sizing rules can be entered for each system. The output of the system
tester includes all sorts of statistics and ratios, including the R-multiple distribution.

Figure 17-7: Trading BloxTM System Editor

Figure 17-8 is an example of how easy it is to use a particular system. It shows the Triple Moving
Average system. It allows you to plug in the size of each moving average, the size of the ATR,
and use different risk amounts for position sizing in a step-like function. You can specify the
range of values and the stepping increment to automatically run a series of tests for all the values
in the specified range. The example in Figure 17-8 will run tests with risk as 0.5%, 1.0%, 1.5%,
2.0%, 2.5%, and 3.0%. It will also vary the size of the stop and the number of days in the ATR
computation across the values specified.

40
The Definitive Guide to Position Sizing Strategies

Figure 17-8: Editing a System in Trading BloxTM

I find the conceptualization of Trading BloxTM to be quite impressive. First, many systems are
included and can be adjusted according to your preference. Second, the more important
components of position sizing and simulation are built into the testing. And it looks easy in that it
just requires mouse clicks and dragging the right bloxTM into the system testing components.
Perhaps this is the future of system testing.

However, let me also be the devil’s advocate. Suppose I simply want to trade something that’s
strongly trending, has a retracement, and then starts to resume its trend. I want to trade that with a
stop below the old retracement. Chances are I probably couldn’t trade it without getting the
version that allows you to do custom programming. And you must ask yourself “Is this what I’ll
probably face?” I don’t have any experience using Trading BloxTM, but this is the sort of issue
that I would guess that most traders will eventually come up against.

And if that’s the case, then you’ll need to learn the “Trading BloxTM Language” and build your
own systems. My impression is that the learning curve for that language might be rather steep, but
that probably depends upon your programming skills.

41
Part II: Position Sizing Software Examined

However, if you want software that requires minimum programming skills, has good systems, and
does position sizing and simulations, then Trading BloxTM Turtle or Trading BloxTM Professional
might be the product you want. For more information, go to their web site, www.tradingblox.com.
The Turtle edition costs $995, the Professional edition costs $1,995, and the Builder edition costs
$2,990.

Wealth-Lab®
Wealth-Lab®13 is often mentioned to me as a platform for doing position sizing. This software is
owned by Fidelity Brokerage; however, the web site says that other data providers are supported,
so the software may be available to people who do not trade through Fidelity. You can apparently
download a demo copy of the software from Fidelity at
http://personal.fidelity.com/products/atp/content/wealthlab.shtml.cvsr.

Wealth-Lab® has a number of features that might appeal to non-programmers:

• Create trading systems just with drag and drop.


• Create indicators and indicators of indicators with drag and drop.
• Test strategies on a whole portfolio with true portfolio level backtesting.
• Optimize a whole portfolio and apply the best values to each symbol with a
single mouse click.
• Apply optimized value on symbol/system combination and on all relevant
tools.
• Real time scanning for systems.
• And Automated Trading Execution via Fidelity.

Of course, my primary interest in doing this review is to focus on the position sizing. My
understanding is that with Wealth-Lab® you can scale in and out of your current position. In
addition, there are a number of pre-programmed position sizing methods. To see the effect of
position sizing you don't need to change the whole code. You just key in a number in the
appropriate field. The software also features (for programmers) the ability to create your own
position sizing strategy (called SimuScript) and apply it to any system on a portfolio level by just
clicking on it. You can create your own performance report metrics as well.

About five pages of the very large manual for Wealth-Lab® are devoted to position sizing. The
basic models are there, but they are very simple as illustrated by Figure 17-9. I scanned through
the entire manual and couldn’t find how to scale in and scale out, even though I was told the
software has the ability to do that.

42
The Definitive Guide to Position Sizing Strategies

Figure 17-9: Wealth-Lab® Position Sizing Screens

However, the SimuScript selection apparently allows you to program all sorts of position sizing
algorithms into the software.

Most of the manual is devoted to the many indicators that are available. And it almost sounds like
the various indicators are thought of as systems. However, I cannot tell that for sure, not being a
member of the Fidelity Brokerage community and never having used Wealth-Lab®.

Here is what one of the reviewers particularly liked about the software:

• “They have done a really good job at trying to break a trading system into
different parts that will allow you to address your trading system in these
different parts.
• For programmers who really want to get control of their trading system, this
is the platform to use.
• They have a drag-n-drop indicator builder that will allow a non-programmer
to build custom indicators by dragging and dropping pre-defined indicators in
a wizard. Then you can custom configure each pre-defined indicator to your
liking.
• I tried TradeStation®, and even took their program training classes and I have
to say that Wealth-Lab® is a much better product for me.”

For more information about Wealth-Lab®, go to http://www.wealth-lab.com. Or if you’d like to


download the trial version and play with it, go to the Fidelity link given earlier.

43
Part II: Position Sizing Software Examined

Mechanica Standard
Bob Spear 14 has released his new software, Mechanica. Mechanica Pro (discussed below) has
been in Beta testing for about four years and it is currently running some serious money for
various CTA firms.

Mechanica Standard is the Windows upgrade to Trading Recipes. Fifteen years ago, Trading
Recipes was definitely my favorite systems development software. However, its dependency
upon the DOS operating system and its position sizing limitations were drawbacks. All of that has
now been fixed by the introduction of Mechanica Standard Edition.

Mechanica starts by building on the foundation of Trading Recipes, and introduces new
functionality in a number of critical areas, such as advanced multi-dimensional trading, and an
enriched programming language that’s easy to use to code and test your trading algorithms and
risk control strategies. If you are not afraid of doing simple programming, its ease of use definitely
makes it worth considering. I found the language in Trading Recipes to be fairly easy and this one
(a superset) is very similar. It has a state-of-the-art electronic help facility and its clickable cross-
reference links are especially useful.

Bob sent me a number of screenshots from the software, but I’ve only included one because I
believe it best illustrates the best feature of Mechanica—the ease of programming your own code.
Figure 17-10 illustrates how simple it is to program a simple position sizing algorithm. It
basically says you have $100,000 in cash and that your position sizing is equal to 2% of equity.
Even I can handle that.

Figure 17-10: Mechanica Rules Editor Showing Easy Programming Language

It’s a unique collection of software tools developed through the years in response to Bob’s own
research and the automation needs of CTAs and hedge fund managers, and has been subjected to
years of real-world testing by others.

44
The Definitive Guide to Position Sizing Strategies

It comes with a number of end-of-day systems built into it, including basic trend following, a pivot
system, a volatility system, a support/resistance system, a parabolic system, and a number of
others. The code for these is already there for you to use or modify.

Mechanica also works with an entire portfolio of positions, not one trade at a time. And it allows
you to work with multiple systems within a portfolio.

You can measure risk, as you choose to define it and view it globally, across every layer of your
portfolio, across multiple systems. You can write your own algorithm for how to define risk and
that will replace the global definition of risk in Mechanica if you tell it to do so. Thus, you could
substitute volatility for risk or any number of other possibilities.

Mechanica has an extensive Monte Carlo suite. For example, you can take the daily percent
changes in your equity curve and do a Monte Carlo analysis to see a range and probability
distribution to show what you might expect in the future based upon the daily changes you have
had in the past. It does NOT do R-multiple simulations.

The software will allow you do most of the position sizing models given in this book. This
includes fixed dollar, fixed percentage, percent risk, percent volatility, group and portfolio heat,
fixed ratio, and to some extent market’s money, scaling in and scaling out. My guess is that some
of these require some programming. More advanced scaling in or scaling out is supported in the
Pro version of the software, which is reviewed later in this chapter. In Bob’s opinion, the need for
these advanced features only comes into play with futures, once your account size reaches about
$1.5 million.

The software also offers portfolio level debugging, accurate forex conversions, has the ability to
run all processes from a batch file, allows you to import an unlimited number of data fields (so
you could do fundamental screening, for instance), automatically nets out commission/slippage
from long and short positions that might happen at the same time in different systems (rather than
charging for them), and it allows you to manage discretionary trades.

Bob has also promised me the ability to do lots of things with R-multiple analysis in the next
version of the software. In addition, although the built-in systems are all end-of-day systems, one
client uses it for intraday analysis though that functionality is not documented or supported.

Generally, if you want software that will allow you to program your own functions in the simplest
way possible, then Mechanica Standard may be the way to go. It costs $3,000 ($995 for an
upgrade) from Trading Recipes and you can get a lot more information at
www.mechanicasoftware.com.

I’ve been using Mechanica in some of the articles in Tharp’s Thoughts (with help from Bob
Spears for the programming), and at this point I’m quite impressed with it.

45
Part II: Position Sizing Software Examined

High End Software (As a Possible Alternative to Building


Your Own)
Many trading businesses simply hire a team of programmers and develop the software that is
needed to do their trading and research. To do this properly, you are probably talking about a
minimum of $250,000 in expenses with no guarantee that you’ll be happy with the software that is
developed. Nevertheless, most professionals seem to go in this direction. For example, at one
point I asked Tom Basso if he thought there was a market for Know Your System among hedge
funds and CTAs. His response was, “I doubt it because they all develop their own software.”
Tom originally developed all of the software that was used to run Trendstat in Foxpro (which later
became Access). However, by the time I met him, he already had a team of programmers on his
payroll.

If you think you have special needs, but have few programming skills and only a moderate budget
(i.e. $30,000 per year) for programming skills, then you may want to consider one of the following
packages.

Mechanica Pro
Mechanica Pro 15 is the first of two advanced software packages that I found. It will be officially
released by the time you read this, but it has been beta tested for many years and, according to Bob
Spear, the developer, it is currently managing significant money for various CTAs. This software,
according to its website, www.mechanicasoftware.com, “Puts you on a level playing field with the
biggest CTAs in the world. It is powerful...”

Mechanica Pro does everything that Mechanica Standard does but offers some additional features,
including 1) Dynamic Risk Management 2) the ability to control multiple accounts 3) the ability to
do options hedging simulations (there are many games that you can play with this) 4) automated
report exporting to Excel so that you can quickly send almost any selection of Mechanica’s
extensive graphical output to Excel and then to a client, 5) custom formatted order sheets and
position output to use for execution and 6) the ability to call Bob and ask questions on issues that
are giving you problems.

The first two features are the real gems. The dynamic risk management basically means that you
change the size of any position in the portfolio at any time based upon what’s happening on any
number of portfolio variables. You can dynamically resize open positions based on any
combination of portfolio-level conditions or events you can envision. This gives you the ability to
research and trade advanced scaling strategies, or make any number of market’s money
adjustments to the portfolio. In my opinion, the ability to do dynamic risk management in a
portfolio with multiple systems and support multiple accounts is an amazing achievement for a
complete trading software package. However, if your account value is less than $500,000, then
dynamic risk management may not be feasible for you to do.

46
The Definitive Guide to Position Sizing Strategies

The second two features really go together. With Mechanica Standard you can do multiple
systems in a portfolio, but if you have multiple clients, each with their own account that you need
to manage, you would need to have a different installation of the software for each account. And
even if you did that, you still couldn’t easily export custom research to a spreadsheet to show
prospective clients, or create custom order sheets for the execution desk. With Mechanica Pro you
can do all of these things.

Mechanica's advanced new multi-account Order Manager is specifically designed for CTAs and
others who manage funds or multiple individual accounts. Featuring custom order and position
reporting, with full batch automation, and advanced account level error detection, Order Manager
helps keep multiple account equity divergence to a minimum.

• For fund managers, order management automation directly translates into administrative
cost reduction, and frees your time for the pursuit of more important matters, like talking to
customers, and increasing assets under management.
• For individual professional traders, it translates into more time spent on other
endeavors...such as research.
• Customization allows you to output order and position reports, formatted to fit your unique
requirements.
• For all traders, Order Manager's error detection helps eliminate potentially expensive order
management snafus.

Bob says, “When a client says, 'Here's a million dollars to trade, what now?' I know exactly what:
Set up the account in Mechanica and push GO, repeat on a daily basis, and watch the system go to
work.

“When the client decides to add funds...Mechanica knows how to rescale the positions and
adjust for the change.” 16

The last feature, I’m not too familiar with, but many funds like to put on an option hedge against a
basket and even define the option pricing model. There are many sophisticated games you could
play doing this and Mechanica Pro allows you to simulate them.

In addition, Mechanica Pro offers you the ability to talk to Bob Spear about any issues or
questions you have. Though custom programming can be done for you on a contract basis, he will
point you in the right direction and give you examples of what you need to do to solve your
problems on your own.

Mechanica Pro is definitely a high-end product, selling for a one time fee of $25,000. The
software also has an optional yearly maintenance fee of $4,500, which is waived the first year, but
gives you access to all of the upgrades and also gives you the free access to Bob Spear.

There is also a System’s Developer Edition of Mechanica. Bob and I didn’t talk about that edition
at all. But if you are a professional who develops trading systems for others, then you might want

47
Part II: Position Sizing Software Examined

to make some inquiries about this edition, which includes security features for protecting your
systems against piracy.

PowerSTTM
PowerSTTM17 is a complete package developed by Bob Bolotin. Bob suggested that people come
to his web site to find out about his product. However, his web site does not discuss the features
of the software. Anyway, with that in mind, here is the description of PowerSTTM that Bob gave
me in a number of emails.

“PowerSTTM is a professional level trading strategy testing product directed towards the more
advanced systems researcher. With a specialty of end of day position trading and portfolio level
money management, PowerSTTM supports integrated portfolio level testing including the ability to
test portfolios composed of multiple markets and multiple systems, advanced portfolio level
money management testing capability, forward trade signal generation, extremely flexible
optimization capabilities, and in general a very powerful and highly customizable testing
environment.

“PowerSTTM is an advanced, highly customizable, highly programmable backtesting software


product with a depth of customization and strategy testing capability.

“For more information please visit the PowerSTTM web site: http://powerst.com/.”

I don’t know the software capabilities, its features, or if there are built in systems. The following
is a quote from an email from Bob Bolotin on its capabilities.

“Something about PowerSTTM is that if a certain type of analysis is not provided, all of this type of
analysis is user programmable (I think I have a good enough idea of what you are getting at in
your list of features to say that with confidence). Most other software would require that the
developer add support for this type of analysis and release a new software version, but that is not
the case with PowerSTTM. PowerSTTM users can program these kinds of concepts themselves.”

Thus it obviously requires programming, and I have no idea what level of skill is required.

“Also, I tend to be customer driven about what features are added to PowerSTTM. If customers
request certain types of analysis I will often volunteer to provide it for them, or at least to help
them get started with sample code (per what I say above that this kind of analysis is end user
programmable). I consider this to be part of the ‘business level’ technical support provided with
PowerSTTM.”

Bob responded to my question in the second paragraph about the level of programming skill
required:

“To answer your question in the above, you are correct that the level of customization I am
referring to requires some programming skill. Programming simple trading systems that are

48
The Definitive Guide to Position Sizing Strategies

relatively simple in other platforms is also relatively simple in PowerSTTM. However, PowerSTTM
also supports programming at a more advanced level, which is what I was referring to in this
paragraph you are asking about.”

The hedge fund manager who developed the R-multiple chart shown on page 260 of Trade Your
Way to Financial Freedom (2nd Edition) uses this software and swears by it.

PowerSTTM costs $25,000 and has a maintenance fee of $1,000 per month, which I assume gives
you a lot of custom help with whatever you need.

My overall impression is that if I were running a trading business and thinking about having
someone develop software for that business, I would certainly look into the option of using either
PowerSTTM or Mechanica.

Conclusion
After writing this chapter, I must admit that I no longer think that the situation involving position
sizing software is dismal. Instead, I think that, depending upon your needs, you can probably find
the software you want. With that in mind, what you purchase will depend upon both your needs
and your skills.

There seems to be plenty of entry and simulation software available. However, I haven’t used any
of it and can’t really offer much of an opinion about any of it. Here, I would recommend that you
go to the web sites, download sample programs or go through the online tutorials, and then decide
what makes sense for your needs. If you go with Mechanica or Trading BloxTM Builder, then you
have simulation software built into your program. I would also guess that you also have this
capability in PowerSTTM, although it is not specifically mentioned anywhere in the web site.

If you have no programming skills and a very low aptitude for programming, then you appear to
have three choices:

• MTPredictorTM
• OmniTrader
• Trading BloxTM (Turtle and Professional Editions)

I would suggest that you check out the possible systems and see if they are something you believe
in and would feel comfortable trading.

If you have some aptitude for programming and want to use the simplest language, then I think
your choice would be Mechanica Standard or Mechanica Pro. Wealth-Lab® might work, although
I don’t know how simple the programming is, however, it is limited to Fidelity Brokerage
Customers. Trading BloxTM Builder might also work here, but my impression is that the language
is a little more difficult.

49
Part II: Position Sizing Software Examined

If your programming skills are good, then you have a wide variety of choices, including
Mechanica, Trading BloxTM Builder, AmiBroker, and PowerSTTM.

Lastly, if you want fairly complete software that will really help you run a trading business, then I
would certainly check out PowerSTTM and Mechanica Pro before looking into custom
programming.

NOTES
1
Ken Long and Leo van Rijswijk reviewed XLQ.
2
Thorsten Reiss filled out a questionnaire on Stator® that I used in this review.
3
Stator® Advanced Finance Management. 2004. Anfield Capital Pty Ltd. 21 Apr. 2007 <http://www.stator-
afm.com/investment-software-purchase.html >.
4
Dave Mabe reviewed StockTickr.
5
Adrian Reid filled out a questionnaire on TradeSim® that I used to write this review.
6
This review was accomplished through a software questionnaire filled out by Thorsten Reiss.
7
Steven O’Keefe filled out a questionnaire that enabled me to do this review.
8
Janeczko, Tomasz. Amibroker. 2001. 26 Apr. 2007 <http://www.amibroker.com/bin/UsersGuide.pdf>.
9
Howard Bandy, Quantitative Trading Systems. Sioux Falls, SD: Blue Owl Press, 2007.
10
Ed Downs filled out a questionnaire on OmniTrader that enabled me to write this review.
11
OmniTrader. 1999. Nirvana Systems. 26 Apr. 2007 <http://omnitrader.com/omnitrader/products/omnitrader.asp>.
12
Curtis Faith filled out a questionnaire that enabled me to write this review.
13
Amanda Tonkin-Hill and Frank Eaves both filled out questionnaires that enabled me to write this review.
14
Bob Spear filled out a questionnaire that enabled me to write this review.
15
Bob Spear filled out a questionnaire that enabled me to write this review.
16
Spear, Bob. Mechanica. 2006. 26 Apr. 2007 <http://www.mechanicasoftware.com/research.htm>.
17
The review was written through numerous emails that I had with the developer, Bob Bolotin.

50